DEVELOPMENT OF A BINDING MODEL TO PROTEIN-TYROSINE KINASES FOR SUBSTITUTED PYRIDO[2,3-D]PYRIMIDINE INHIBITORS

Previously, our laboratories have reported on a new class of highly po tent tyrosine kinase inhibitors based on the pyrido[2,3-d]pyrimidine c ore template. To understand the structural basis for the potency and s pecificity, a model for the binding mode of this class of inhibitors t o the tyrosine kinase domains of c-Src, PDGFr, FGFr, and EGFr tyrosine kinases was developed from structural information (principally utiliz ing the catalytic domain of c-AMP-dependent protein kinase as template ) and structure-activity relationship (SAR) information. In the result ing docking mode, the pyrido[2,3-d]pyrimidine template shows a hydroge n-bonding pattern identical to that of olomoucine. The 8-aryl substitu ent of the heterocycle is located deep in the binding cleft in a pocke t not used by ATP, which helps to confer high-affinity binding as well as specificity. The 2-anilino and 2-(dialkylamino)alkylamino substitu ents as well as the 7-urea substituent of inhibitors within this class are located at the entrance of the binding cleft and make contact wit h residues in the hinge region between the two kinase lobes. This allo ws considerable variability and bulk tolerance for C-2 and N-7 substit uents. The models presented here are consistent with the SAR seen for the inhibition of a number of isolated enzymes and provide a structura l basis to explain their specificity. They have been used successfully to design new highly potent protein kinase inhibitors.